TW201351416A - Method for improving performance when flash memory storage device works in wide temperature range - Google Patents

Abstract

A method of a flash memory storage device using Read Retry method is disclosed. This method includes using a thermal sensor to records temperature information while programming flash memory, and using this temperature information to compensate the temperature difference between program and read operations to improve Read Retry performance.

Description

Method for improving the characteristics of flash memory storage devices operating over a wide temperature range

The present invention relates to improving the characteristics of flash memory, and more particularly to a method for reducing the number of tests in a read retry operation, and the less the number of tests used in the read retry operation, the less flash memory storage The features of the device are better.

Currently, flash memory is very common in storage systems. Different types of memory technologies use different flash types. The NAND flash memory system is used to store one of the most popular memory devices. Because of its fast speed, high density, low power consumption and low cost, NAND flash memory system is widely used in mobile systems, including mobile phones, MP3 players, digital cameras, and portable devices. Computer and so on. For mobile systems, temperature tolerance is an important factor. Temperature sensitivity is a feature of NAND flash memory, especially for more advanced processes. When an error bit number exceeds the correction limit capacity of Error Correction Coding (ECC), the Read Retry Method is used to correct this problem. The current read retry method is able to overcome this problem, but it is usually not effective. From the largest to the smallest test of different sense voltages, the current read retry mode can take some time before finding a suitable sense voltage value. This read retry sequence can result in significant burst performance degradation. The stable read/write feature is important for mobile applications.

Please refer to FIG. 1, which is a flow chart showing a conventional flash memory write command process.

The steps of the conventional flash memory write command process are as follows: step S11: receiving a "write" command (CMD); and step S12: writing data to a flash memory.

Please refer to FIG. 2, which is a flow chart showing a conventional flash memory read command process.

The step of the conventional flash memory read command process includes: step S21: receiving a "read" command (CMD); step S22: fetching data from a NAND flash memory map; step S23: Determining whether the Cyclic Codes are met; if so, stopping the program; if not, proceeding to the next step; Step S24: Confirming whether the error correction coding (ECC) engine can be corrected; if so, correcting the error The encoding engine starts to make corrections (step S241); if not, proceeds to the next step; step S25: provides a voltage VT1 as a first test voltage for reading retry to a memory cell array (memory cell array) Step S26: reading data from the NAND cache memory; step S27: determining whether the cyclic code is consistent; if yes, performing steps S241; if not, proceed to the next step; step S28: confirm whether the error corrects the encoding engine can be corrected; if yes, execute step S241; if not, proceed to the next step; and step 29: confirm whether all is available The test voltage VTn has been tried; if so, an uncorrectable error is made (step S291); if not, another test voltage VTn is supplied (step S292) and the process returns to step S22.

When an error bit number exceeds the capacity of the error correction coding (ECC) engine, a plurality of sensing voltages are sequentially tested. These test voltages start from the lowest voltage level voltage VT1 and the highest voltage level voltage VTn. Once the number of error bits in the flash memory data is covered by the error correction coding (ECC) engine, the error correction coding (ECC) engine will recover these error bits.

However, as shown in FIG. 3, the first pulse curve CU1 at a first temperature such as room temperature and the second pulse curve CU2 at a second temperature such as a programmed temperature higher than room temperature are Different, where the horizontal axis represents time and the vertical axis represents data volume. At the first temperature, the process is from voltage VT1 to voltage VTn, such as voltage VT1 to voltage VT7. However, at the second temperature, the second pulse curve CU2 is shifted to the right of the first pulse curve CU1, and the program is also from the voltage VT1 to the voltage VTn, for example, compared with the first pulse curve CU1, the voltage VT1 to the voltage VT2 is Failed and voltage VT3 to voltage VT7 It is passed.

Regardless of the temperature, the program always starts from voltage VT1, making the test waste too much time.

Please refer to U.S. Patent Publication No. 20100322007 (hereinafter referred to as '007 patent), which discloses a flash memory for reading data and a method thereof. The method includes performing a test read operation directed to the test data, the test data being stored in a memory cell array of the flash memory in a sequence of test read retry operations. , wherein each successive test read retry operation uses a test read voltage level that is higher than a previous test read retry operation, until one of the tests in the sequence test read retry operation Reading a retry operation, successfully reading test data using a minimum test read retry voltage associated with the one test read retry operation; setting an initial read voltage for the flash memory device, It is equal to the minimum read retry voltage, and then performs a standard read operation, which is directed to the user data, which is stored in the memory cell by repeatedly providing a sequence of read retry operations. In the array.

However, the method disclosed in the '007 patent corrects errors from flash memory at different temperatures, but is less effective.

Based on the above problems, the inventors have proposed a method of improving the characteristics of a flash memory storage device operating over a wide temperature range to overcome the deficiencies of the prior art.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a method of improving the characteristics of a flash memory storage device operating over a wide temperature range by reducing the test time based on the current temperature of a temperature sensor to improve efficiency.

To achieve the above object, the present invention provides a method for improving the characteristics of a flash memory storage device operating over a wide temperature range, the steps comprising: step SA01: receiving a read command; step SA02: fetching memory from a NAND Reading data; step SA03: judging whether the cyclic code is consistent; if yes, ending the process; if not, proceeding to the next step; step SA04: confirming an error correction encoding engine can make correction; if so, the error Correcting the encoding engine begins to make corrections; if not, proceeding to the next step; step SA04': obtaining a current temperature and a stylized temperature from a temperature sensor; step SA05: confirming a read retry record Confirming whether a physical address has a read retry mode and having an appropriate initial test voltage record; if so, skip to the next two steps; if not, proceed to the next step; Step SA06: Calculate Retrieving the initial test voltage according to the current temperature and the programmed temperature; step SA07: providing the read retry start test Pressing a memory cell array; step SA08: reading data from the NAND flash memory; step SA09: determining whether the cyclic code is consistent; if so, storing the current test voltage to the read retry record; Otherwise, proceed to the next step; Step SA10: confirming whether the error corrects whether the encoding engine can make correction; if yes, performing step SA04; if not, proceeding to the next step; and step SA11: confirming whether all available test voltages have been tried; If so, a report that the error cannot be corrected; if not, provide another test voltage and return to step SA08.

Wherein, the temperature sensor is disposed inside or outside the NAND flash memory.

While the invention has been described in terms of several preferred embodiments, the preferred embodiments of the present invention It is not intended that the invention be limited to the following drawings and embodiments.

Please refer to FIG. 4, which is a flow chart showing a flash memory executing a write command according to the present invention.

A flow of executing a write command by a flash memory according to an embodiment of the invention includes the steps of: step S31: receiving a write command; step S32: obtaining a current temperature from a temperature sensor; and step S33: recording the current temperature As a stylized temperature (TEMP_prog); and step S34: writing data to a flash memory.

The above temperature sensor (not shown) may be disposed inside or outside the flash memory, but is not limited thereto.

Please refer to FIG. 5, which is a detailed flowchart showing a flash memory executing a read command according to the present invention.

A detailed flow of executing a read command by a flash memory according to the present invention includes the steps of: step S401: receiving a read command; step S402: reading data from a NAND flash memory; and step S403: determining whether the cyclic code is Alignment; if yes, end the process; if not, proceed to the next step; step S404: confirm an error correction whether the encoding engine can make corrections; if so, the error correction encoding engine begins to correct (step S414); If not, proceed to the next step; step S404': obtain a current temperature and a programmed temperature (TEMP_prog) from a temperature sensor disposed inside or outside the flash memory, wherein the current temperature system When the flash memory is not used or is sensed by the temperature sensor before being programmed, the programmed temperature is sensed by the temperature sensor when the flash memory is being programmed; step S405: confirming a read Retry the record list; that is, confirm whether a physical address has a read retry method and has an appropriate initial test voltage record; , Then jumps to step S407; if not, then proceed to the next step; step S406: According to the current temperature and the temperature of the stylized read retry starting to calculate the test voltage; Step S407: providing the read retry start test voltage VT to a memory cell array; step S408: reading data from the NAND flash memory; step S409: determining whether the cyclic code is consistent; if yes, the current test voltage VT is stored in the read retry record table (step S413); if not, proceed to the next step; step S410: confirming whether the error correction encoder can make corrections; if so, then skipping to step S414; if not And proceeding to the next step; and step S411: confirming whether all available test voltages VT have been tried; if so, presenting an uncorrectable error report (step S415); if not, providing another test voltage VT The process returns to step S408 (step S412).

In step S401, a flash controller receives the read command from a host (HOST). In step S402, the flash controller translates the logical address into a physical address and reads the data from the NAND flash memory. In step S403, the flash controller uses the data read from the NAND flash memory to calculate the cyclic code and compares it with the cyclic code stored in the NAND flash memory. In step S404, the error correction encoding engine confirms whether the number of error bits is within the maximum error bit tolerance range. In step S406, the flash controller confirms the programmed temperature of the desired material and compares it with the current temperature. Using some arithmetic, the flash controller uses a proper value for the initial test voltage VT. In step S408, the flash controller is Reading data from NAND flash memory. In step S409, the flash controller uses the data read from the NAND flash memory to calculate a cyclic code and compares it with the cyclic code stored in the NAND flash memory. In step S410, the error correction encoding engine confirms whether the number of error bits is within the maximum error bit tolerance range. In step S411, the number of test voltages VT can be predefined. In step S412, a test voltage VT is provided according to the current temperature and the programmed temperature (TEMP_prog). The voltage value of the next test voltage VT can be determined by some arithmetic of the current temperature and the programmed temperature. In step S414, the error correction encoding engine is executing and obtaining the correct material. In step S415, the data uncorrectable information is returned.

The temperature information is obtained from a temperature sensor and is recorded before the write command is completed. This temperature information represents the stylized temperature of the data. When this data is requested by the host (HOST), its reading process is shown in Figure 5. If the temperature at this time is very different from the temperature at the time of stylization, the distribution will be displaced by a distance. It is the temperature difference between stylization and reading, such as T-factor. With the T-factor information, the offset voltage has been selected to select the initial test voltage. It means that some non-essential test voltage levels can be skipped in advance. The next test voltage is provided if the number of error bits with the sense voltage known to read the data exceeds the error correction code engine capacity. Since there is an additional T-factor, the direction of the next test voltage can be determined. That is, the next test voltage can be one of a higher voltage level or a lower voltage level. The test voltage is recorded once a sense voltage is obtained with data that corrects the number of error bits. If the next page is read by the host, this information will be used to define the initial test. The test voltage is used as a reference.

Referring to FIG. 6, an illustration of reading retry in FIG. 5 is shown.

a first pulse curve CU1' at a first temperature system such as room temperature and a second pulse curve CU2' at a second temperature higher than room temperature, wherein the horizontal axis is time, The vertical axis is the amount of data. At the first temperature, the flow is from test voltage VT1 to VTn, such as VT1 to VT7. Then, at the second temperature, the second pulse curve CU2' is shifted to the right of the first pulse curve CU1' by a T-factor distance, but the flow only starts from the test voltage VT3 to the test voltage VTn, for example, A pulse curve CU1' and FIG. 3 are compared, the test voltages VT1~VT4 can be skipped directly, and the test voltages VT5~VT7 are qualified.

For example, a user takes a picture in the Arctic. It means that the data is programmed into a flash memory at minus 20 degrees Celsius. After the trip, the user returned to the tropics and showed the photo to a friend, that is, the data was read from the same flash memory at 40 degrees C. A temperature difference of more than 60 degrees can cause a voltage shift as shown in FIG. For the conventional read retry method, it is required to test from the test voltage VT1 to the test voltage VT5 in each loop, and the correct data has been obtained. For the reading method of the present invention, it is possible to start the test directly from the test voltage VT5 without additional testing, and at least improve the test efficiency by 30%.

Therefore, the reading method according to the present invention can shorten the test time according to the current temperature sensed by a temperature sensor, thereby improving the test efficiency.

Although the present invention has been explained in connection with the preferred embodiments, it is not intended to limit the invention. It should be noted that those skilled in the art according to the present invention The idea can be constructed in many other similar embodiments, all of which are within the scope of the invention.

[this invention]

CU1‧‧‧ first pulse curve

CU1’‧‧‧ first pulse curve

CU2‧‧‧ second pulse curve

CU2’‧‧‧second pulse curve

TEMP_prog‧‧‧ Stylized temperature

VT‧‧‧ test voltage

VT1~VT7‧‧‧ test voltage

VTn‧‧‧ test voltage

Figure 1 is a flow chart showing a conventional flash memory write command process.

2 is a flow chart showing a conventional flash memory read command process.

Figure 3 is a schematic diagram showing the read retry in Figure 2.

Figure 4 is a flow chart showing the execution of a write command by a flash memory of the present invention.

Figure 5 is a detailed flow chart showing the execution of a read command by a flash memory of the present invention.

Fig. 6 is a view showing an example of reading retry in Fig. 5.

Claims (10)

A method for improving the characteristics of a flash memory storage device operating over a wide temperature range, the steps comprising: step SA01: receiving a read command; step SA02: reading data from a NAND cache memory; step SA03: determining a cyclic code Whether it is met; if yes, end the process; if not, proceed to the next step; Step SA04: Confirm an error correction code engine can make corrections; if so, the error correction coding engine begins to correct; if not If yes, proceed to the next step; step SA04': obtain a current temperature and a stylized temperature from a temperature sensor; step SA05: confirm a read retry record table; confirm whether a physical address has been provided a read retry mode with an appropriate initial test voltage record; if so, skip to the next two steps; if not, proceed to the next step; step SA06: calculate the current temperature and the programmed temperature a read retry start test voltage; step SA07: providing the read retry start test voltage to a memory cell array; step SA08 : reading data from the NAND flash memory; step SA09: determining whether the cyclic code is in compliance; if so, storing the current test voltage to the read retry record table; if not, continuing Going to the next step; step SA10: confirming whether the error corrects the encoding engine for correction; if so, performing step SA04; if not, proceeding to the next step; and step SA11: confirming all available test voltages Has it been tried; if so, a report that cannot correct the error; if not, provide another test voltage and return to step SA08. The method of claim 1, wherein the temperature sensor is disposed inside or outside the NAND flash memory. The method of claim 1, wherein in the step SA01, a flash controller receives the read command from a host (HOST). The method of claim 3, wherein in the step SA02, the flash controller translates a logical address into a physical address and flashes the memory from the NAND Read the data. According to the method of claim 4, wherein in the step SA03, the flash controller uses the data read from the NAND flash memory to calculate the cyclic code and stores it in the NAND. The cyclic codes in the flash memory are compared. The method according to claim 5, wherein in the step SA04, the error correction coding engine determines whether the number of the error bits is within a maximum error bit tolerance range. According to the method of claim 6, wherein in the step SA06, the flash controller confirms the programmed temperature of the required data, and is currently The temperature is compared, wherein the flash controller uses an appropriate value obtained by an arithmetic to give an initial test voltage. According to the method of claim 7, wherein in the step SA08, the flash controller reads data from the NAND flash memory, and in the step SA09, the flash controller is The data read from the NAND flash memory is used to calculate each of the cyclic codes and compared to the cyclic code stored in the NAND flash memory. According to the method of claim 8, wherein in the step SA10, the error correction coding engine determines whether the number of the error bits is within the maximum error bit tolerance range, in the step SA11. The number of such test voltages is predefined. According to the method of claim 9, wherein in the step of providing another test voltage and returning to step S408, a test voltage is provided according to the current temperature and the programmed temperature, the next one The voltage value of the test voltage is determined by an arithmetic of the current temperature and the programmed temperature. In the step of correcting the error correction engine starting the correction, the error correction coding engine is executed and the correct data is obtained. In this step of the report of the uncorrectable error, the information of the error that the data cannot be corrected is returned.